Printer Friendly

Planar millimeter-wave arrays.

Planar Millimeter-Wave Arrays (*)


The applications for planar mm-wave antennas have expanded over the years for both military and commercial markets. Typical military uses include smart munitions, missile guidance and fuzing, as well as peacetime military applications, such as RPV surveillance. Commercial uses include applications such as collision avoidance, perimeter security surveillance and remote sensing.

This is a relatively mature field. Early work in microstrip planar arrays [1] led to subsequent application of this planar technology to mm-waves as early as 1976. Microstrip series arrays described in 1979 [2] have been applied in antennas up to 94 GHz as a way to reduce coupling, radiation and loss associated with corporate feed lines. In addition, the frequency scanning properties of series arrays may be used for steering a mm-wave antenna. The advantage of printing monopulse circuitry and array quadrants on the same layer, thereby reducing the cost and complexity of interconnections and phase matching, is highly attractive at mm-wave frequencies. A particularly important architecture is the incorporation of active components into a compact antenna system. This potentially would reduce complex interconnections between sybsystems and result in a more reliable and affordable system.

The pros and cons of planar mm-wave systems are well-known, and trade-offs must be studied when considering a planar antenna for mm-wave application. On the pro side, the planar antenna is thin, compact and lends itself to high volume production at low cost. However, one must consider circuit losses, transmission line radiation and coupling between lines for a dual polarized antenna.

This paper describes three concepts that are in the development process and will be used for the development of a dual polarized, low sidelobe, dual axis monopulse, planar mm-wave antenna. These three concepts are a compact dual polarized element, an array architecture with a center hole and a dual sided feed network for a dual polarized array.

Radiating Element

Dual polarized antennas are useful in communication and polarimetric radar applications. Existing technology for dual polarized radiators in a plane include a dual polarized patch radiator, a dual of a patch radiator, crossed dipoles, crossed slots and a dual polarized waveguide. The dual polarized patch radiator is made possible by exciting two orthogonal modes in a square or circular metallic patch. A schematic diagram of this radiator is shown in Figure 1. A dual of a patch radiator uses the same principle as the dual polarized patch radiator with the dual of a microstrip patch. Crossed dipoles consisting of two dipoles arranged in an orthogonal orientation will produce dual linear polarizations with collocated phase centers, as shown in Figure 2. Crossed slots are the dual of the crossed dipole configuration and produce dual linear polarization with collocated phase centers. A dual polarized waveguide is two orthogonal modes that are excited in the waveguide. This is a more efficient element than the printed circuit elements; however, it requires a three-dimensional cavity, as shown in Figure 3. Other dual polarized elements, such as the dual polarized sinuousoidal antenna, are not suitable at mm-wave frequencies.

All of these elements exhibit acceptable dual polarized performance. By exciting dual orthogonal modes on the radiator, they eliminate the need for an external mode transducer. In theory, these elements have a large cross polarization rejection. In practice, however, coupling between modes will produce a cross polarization component at larger than acceptable levels. In addition, the resonant size of these elements creates feedline crowding that increases cross polarization coupling. Techniques have been developed and implemented to increase isolation between polarizations. Figure 4 shows one solution of separating the polarizations into two apertures. One disadvantage is the 3 dB loss in directivity. However, this may be outweighed by the advantage of increased cross polarization rejection, increased port-to-port isolation and decreased line losses. One technique to increase cross polarization rejection within the same aperture is implemented in the antenna shown in Figure 5. Pairs of microstrip elements are fed in phase for one polarization and out of phase for the orthogonal polarization. This technique was patented [3] and has been implemented in several planar arrays, including applications at mm-wave frequencies. Using this technique, cross-polarization components may be reduced to levels of -25 dB in the main beam. In addition, port-to-port isolation levels of -30 dB are achievable. However, implementing this technique may lead to higher grating lobes and associated losses that may not be acceptable for systems seeking a low sidelobe radiation pattern.

Dual Polarized Annular Ring

This element has several advantages that reduce many of the existing problems in present technologies. A resonant annular ring is etched in the ground plane. The element is fed by a microstrip line printed on the opposite side of the substrate material. The slot is a bidirectional radiator and, therefore, requires a reflective cavity back. Several attractive properties of this radiator, including small size, open microstrip feed line, dual size feed, broad single element radiation pattern, variable geometry, dual polarized aperture, additional feed network and broader bandwidth, make it well suited for planar mm-wave antennas.

The resonant circumference of the slot is one wavelength long. Thus, the diameter is [lambda]/[pi]. Circular patch radiators, dipoles and slots have a dimension on the order of 0.5[lambda]. The smaller size has several advantages, including closer array spacing and/or decreased feedline crowding. The open microstrip feed line enables the antenna to be manufactured without plated through via holes. With the dual sided feed, the element can be fed from opposite sides of a laminate, allowing feeds for each polarization to be separated and isolated. The broad single element radiation pattern is similar to the broad E-plane pattern of slot and patch radiators with a small cross polarized component due to fields radiating from opposite sides, as shown in Figure 6. Variable geometry is not necessary for the element to be implemented in a circular shape. The element may be a square or a diamond with a perimeter of one wavelength. Having a dual polarized aperture is an important property of the radiator because it has a null located 90[degrees] from the feed point, enabling its use as a dual polarized element. With an additional feed network, the element can be fed with 0/180[degrees] network from opposite sides of the laminate and enhances the performance characteristics of the element in three ways. The network adds symmetry to the E-plane pattern; it increases the peak directivity of the element; and it increases the isolation between polarizations. The dual polarized microstrip patch radiator has inherent bandwidth problems. However, is directly proportional to the thickness of the substrate. However, as the substrate thickness is increased, feedline radiation increases, causing increased losses and radiation pattern degradation. The bandwidth for low cost, mass production microstrip planar mm-wave antennas is less than 5 percent of the center frequency. This limited bandwidth makes the resonant frequency of the patch resonator susceptible to variations in the dielectric constant. The bandwidth of the dual polarized annular slot is on the order of 15 percent, and is, therefore, much less susceptible to changes in the dielectric constant.

Array Layout

There are several reasons why an array layout with the center cut out is attractive from the systems standpoint. One reason is the desire to place the monopulse comparator on the printed circuit board as a part of the monolithic feed for each polarization. In order to accomplish this, the array needs to be thinned in the center where the feedlines to each quadrant are accessible. Another reason is that it is often desirable for another independent sensor to be integrated with the antenna. This is true in the case of smart munitions, where it is advantageous to have both a mm-wave sensor and an IR sensor in the forward section of a smart projectile. Depending on the size of the hole in the center of the array, either or both of these functions can be performed.

Altering the array architecture to accommodate either of the above options adversely affects the performance. One problem is the disruption of the aperture's low sidelobe distribution. When a weighting function is placed across the aperture, the center elements are most heavily weighted. By removing these center elements, the sidelobe performance degrades by a nonlinear function related to the size of the hole. The second problem is the obvious decrease in directivity of the array pattern. Figure 7 is a graph of the increase in sidelobe levels and the decrease in directivity vs. the size of the center hole in wavelengths for an aperture of 16[lambda] in diameter. Figure 8 shows a circular array with a 30 dB Taylor distribution. Figure 9 shows the pattern of the same aperture with the array thinned in the center superimposed onto the nonthinned pattern. This direct comparison clearly illustrates the thinning effects. In this case, one must trade-off the size of the hole necessary to include an onboard monopulse comparator and/or IR sensor and the pattern degradation acceptable to the system. In addition to the size of the center hole, the array architecture has to be considered. In creating an area in the center of the array for the monopulse circuit, the option of nonsymmetric planes may be available. In creating an open area for an IR sensor, this option does not exist. The power combining technique and phase length of feed lines must be taken into consideration to determine the element symmetry within the quadrant. The aperture amplitude distribution has to be evaluated in terms of the feasibility of the power splits in the power combiner/divider. Thinning must be accomplished with the entire array architecture in mind.

Low sidelobes for a monopulse antenna in the sum radiation pattern only may not be sufficient. In a monopulse tracking system, low sidelobes are necessary in both the sum and difference modes in order to reduce the antenna's susceptibility to jamming threats. This creates the additional design problem of achieving simultaneously low sidelobes in the sum and the difference radiation patterns. The Taylor and the Bayliss aperture distribution differ in both amplitude and phase. It is easy to see from Figure 10 that attaining low sidelobes in both sum and difference from a single feed network would be difficult. In order to use both a Taylor and a Bayliss distribution in the same aperture, two separate feed networks plus a switch between them would be necessary. Many system designers will take the difference sidelobe degradation of the single amplitude distribution in order to retain a monolithic design with low sidelobe sum radiation patterns. Conventional monopulse antennas with a single amplitude distribution simply apply the Taylor distribution across the aperture and achieve low sidelobes in the sum radiation pattern and high sidelobes in the difference radiation pattern.

One key innovation for solving the simultaneously low sidelobe problem is to break up the aperture into several sections. The sum radiation pattern is created by illuminating all five sections with a low sidelobe distribution. The difference radiation pattern is created by using only the four outer sections of this distribution. Using this method, the sum sidelobes are optimized and the difference sidelobes are greatly reduced over conventional monolithic methods. An antenna using this improved monopulse design technology is shown in Figure 11.

This technology does solve the single distribution problem for a full aperture. The case of the center thinned aperture also can be optimized to achieve low sidelobes in both sum and difference modes. Although the sidelobe levels in the sum distribution are a compromise from an unthinned array, the system sidelobes are reduced greatly over the conventional monolithic monopulse approaches. Theoretical radiation patterns generated from measured data are shown in Figure 12. The sidelobes in both sum and difference modes are 18 dB down from the main lobe.

Dual Polarized Feed Network

The feed network for a large planar array is the largest contributor to radiation pattern distortion. This problem becomes worse when a second feed network is introduced as in the case of a dual polarized array. Feedline crowding occurs, causing line-to-line coupling, line-to-patch radiator coupling and feedline radiation. Line-to-line coupling creates several problems in the array. Coupling that occurs between feedlines of opposite polarizations reduces the isolation between polarizations. Coupling between lines of the same feed network can cause impedance mismatches in the junctions and disrupt the phase and amplitude distribution of the aperture. This results in a higher sidelobe level than the theoretical design predicted. There are several methods for solving this problem.

Narrowing the line widths is one way to increase usable real estate and reduce coupling. This can be accomplished by designing a higher impedance feed network, which results in greater feed line losses. The thickness of the board material can be reduced to narrow the line widths. This also produces greater feed network losses. In addition, as the thickness of the board decreases, the bandwidth of the antenna element also will be decreased, making the antenna array less producible due to its sensitivity to variations in dielectric constant.

An array designer often will increase the spacing between elements in the array to accommodate dual feed networks. This is an unacceptable solution in a low sidelobe array due to the increase in grating lobes. Several designs use series fed arrays to decrease feedline overcrowding. However, series fed arrays are dispersive. Beam scanning will occur with changes in frequency or changes in dielectric constant due to temperature. Depending on the size of the array, this can be detrimental to beam pointing accuracy.

Using the dual polarized annular slot element permits separation of the feed networks of the two polarizations. The slots are etched into the ground planes of two separate feed boards. These feed boards then are laminated with ground planes together and slot positions aligned. In this manner, the feed networks can be separated while maintaining a shared aperture. This increases the real estate available for the feed networks of the two polarizations. The designer is now able to implement a nondispersive, low sidelobe corporate feed network for each polarization.

Final Array Design

These three concepts have been integrated into the array currently being developed. Figure 13 shows a photograph of both sides of the mm-wave planar array. The feed network has been further simplified by quantizing the amplitude distribution into four element subarrays. The power combiner has been measured for phase and amplitude distribution. Table 1 shows the deviation from expected results. Figure 14 is a comparison of theoretical and expected far field radiation patterns generated using this measured distribution.


[1] R.E. Munson, "Conformal Microstrip Antennas and Microstrip Phased

Arrays," IEEE Trans. AP-S, Vol. 22, No. 1, January 1974, pp. 74-78.

[2] T. Metzler, "Microstrip Series Arrays," Proc. Workshop Printed Circuit Antenna Tech., October 17-19, 1979, New Mexico State University, Las Cruces, NM, pp. 174-178.

[3] Patent #4,464,663 by F. Lalezari and R. Munson, "Dual Polarized High Frequency Microstrip Antenna."

Farzin Lalezari received his BSEE degree, summa cum laude in 1976, and his MSEE in 1977 from Polytechnic Institute of Brooklyn. He continued his graduate work at the University of Colorado in 1978 and 1979. Since 1979, he has been at Ball Aerospace Systems Group, where he has been in charge of the development of more than 50 antenna designs. Currently, he is manager of passive arrays at Ball, Communication Systems Division.

Theresa C. Boone received her BSEE degree from the University of Rochester, Rochester, NY, in May 1985. She has done graduate work in the field of electro-magnetic at the University of Colorado, Boulder. Since June 1985, Boone has been with Ball, Communication Systems Division. She is an antenna design engineer, responsible for the development of numerous frequencies. Currently, she is working on several broadband, dual polarized, low loss phased arrays.

J. Mark Rogers received his associate's degree in science in 1990 from Front Range Community College, Westminster, CO. Currently, he is with Ball, Communication Systems Division, where he is employed as an antenna designer. Rogers has seven years experience in design, fabrication and testing of antenna systems.

(*) Invited paper
COPYRIGHT 1991 Horizon House Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:antennas
Author:Lalezari, Farzin; Boone, Theresa C.; Rogers, J. Mark
Publication:Microwave Journal
Date:Apr 1, 1991
Previous Article:MDHL: bridging the design and communication gap.
Next Article:Broadband characterization of conductor-backed coplanar waveguide using accurate on-wafer measurement techniques.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters